Herd Immunity III
Double Dip
A small update to my herd immunity tracker. The main change is that the data now accounts for the fact that people previously infected are getting vaccinated. I assume that past infection and vaccination are uncorrelated. This means that if, on one day a state has an estimated total infection rate of 30%, 30% of the people vaccinated on that day will have previously had the virus. The '% Infected and Vaccinated' data only refers to people who were infected then vaccinated, not vaccinated and contracted the virus. I'm not yet modeling those who get infected after vaccination.
Infection and vaccination are not happening to random samples of the population, but I think assuming the groups are uncorrelated is a good first approximation. The main groups currently receiving the vaccine are healthcare workers, essential workers, seniors, and those with preexisting conditions. While the latter groups have likely been taking higher than average precautions (and thus have lower infection rates than the overall population), healthcare and other essential workers are some of the most exposed to the virus and will thus have higher than average infection rates.
As of February 7th, accounting for previous infection among the vaccinated lowers the estimate of immunity from 37.2% to 34.8% of the US population.
The other change is a slider to make it possible to examine the data over smaller time periods. By default, the window is set from the first day vaccines were available in the US (December 14th, 2020) to the most recent day of data, but you can use the slider to change the window to look back before the vaccine rollout or a shorter period after rollout.